Why do field-defect counts and failed-row counts differ?
One record can fail several checks at once. Adding role, company and address defects therefore overstates the number of records needing attention when the categories overlap. Keep a field-level defect count and a separate unique-row disposition, so the repair workload and usable inventory are both visible.
Apollo’s search documentation includes job-change-related email status information. That is a reminder that role changes and address status are distinct signals. It is not evidence of a universal decay rate, and it does not tell you what share of your own list has changed.
Start with the decision your campaign needs. If the offer is for operations directors at independent retailers in a particular territory, the role, ownership and location matter alongside the email. A technically reachable former director can still waste qualification time. Save the original claim and the new evidence for each reviewed field.
Sources: Apollo prospect search filters
How should a field-level freshness audit work?
Freeze a defined cohort and inspect the fields required by its campaign. Record the evidence date and a separate result for each field, allowing unknown values. Calculate field failure rates with visible denominators. Use an exclusive overall disposition when reporting rejected records so overlapping defects do not inflate the total.
Illustrative audit: among 100 randomly selected records, 12 have a role mismatch, eight have a company-fit mismatch, and five have an address failure. These categories overlap: suppose 20 distinct records have at least one defect. Field counts total 25, but the record-level rejection rate is 20%, not 25%.
If a role repair costs six minutes and an address repair costs two, prioritize by the campaign’s required fields and the value of recovering a record. Do not spend on a better address for an account that has already failed the audience rule. The costs and counts here are assumptions for planning, not field observations.
| Check | Records failing | Interpretation |
|---|---|---|
| Role | 12 | Wrong decision-maker for this brief |
| Company fit | 8 | Account outside the target |
| Address | 5 | Required channel failed review |
| Any defect | 20 distinct | 20% fail at least one check |
Which records deserve a refresh first?
Refresh records whose required facts are uncertain and whose account fit still justifies the work. Review evidence age alongside observed changes, not as a universal expiration date. Keep rejected accounts out of the repair queue. A recent check is useful evidence about its scope, not a permanent guarantee of accuracy.
Build a queue with the failed field, evidence source, last checked date, owner and next action. Separate a recoverable missing title from a confirmed account mismatch. Review the highest-value, time-sensitive accounts first, while maintaining a reproducible sampling method for the rest of the cohort.
For a fresh research order, specify which fields must be current and how they will be evaluated. HEED’s terms define accepted record fields and review responsibilities. A saved Last Checked date should remain attached to its actual verification event; copying or importing a row does not refresh it.
- Do not replace a field’s evidence date with the export date.
- Keep unknown values visible rather than treating them as valid.
- Record unique rejected rows separately from the sum of field defects.
Sources: HEED service and record acceptance terms
Define a B2B research briefSpecify the audience, required fields and exclusions before choosing a batch.
Common questions
Does a non-bounce prove the contact is still in the right job?
No. Address acceptance and job fit are different checks. A role may change while the mailbox still receives mail or forwards it.
Should every record expire after a fixed number of days?
There is no universal shelf life established by this model. Use your audience, field volatility, evidence and measured audit results to choose a review interval.